-
Notifications
You must be signed in to change notification settings - Fork 0
Home
Andreas Høyer Iversen edited this page Mar 8, 2015
·
5 revisions
This project is an experiment to see if an AI can learn to learn to play chess. That is, can a program learn to play chess without any pre-coded knowledge about chess? The idea is to let the program try to play against another opponent and incrementally get better by learning from mistake. One of the great challenges with this is the required ability to think multiple steps ahead without being explicitly told to do so. On the benchmark site is some of results achieved until now.